Strategy Guide for RNA Lab

  • 16
  • Article
  • Updated 6 years ago
  • (Edited)
mpb21's Strategy Guide for Solving RNA Lab Structures.

If you are new to the game, I recommend reading the strategy guide for solving the challenges:(http://getsatisfaction.com/eternagame/topics/the_strategy_guide_to_solve_eterna_puzzles).

Then play through several of the challenges until you get a feel for the controls and you start to develop some intuition for the process.

Many of the strategies used in the strategy guide will be used here, but not explicitly mentioned, so if you haven't yet, read both.

The main difference is that for the challenge puzzles, you are only looking for one correct solution to fulfill the requirements for the puzzle (structure, number of bonds etc...). However in the lab puzzles, you are looking for the best solution. This is a crucial distinction that you must be aware of, because currently, for each round, only one (if any) of your solutions will be synthesized. So you need to make sure that it is the best possible.

Something to remember when working in the RNA Lab:
-The scoring function in EteRNA which tells you whether a given sequence will or not is good, but it's far from perfect.
-So you want to make sure that the sequence that you submit will be safely on the side of folded. This is where steps 3 and 4 (below) come in handy.

Steps for solving:

1.
Filling In. Start as you would for a challenge puzzle by filling in U's to pair with the A's that are there initially. You will want to do this in randomly alternating segments of 1-3.

2.
Structure Stabilization. Then fill in G-C pairs at the base pairs that are adjacent to the loops. Use a minimal amount more if needed to stabilize the structure. In theory, you could submit your score at this point, but there are a few extra things that you can do to make your structure just that much better.

3.
Loop Stabilization. The loops of the RNA are the most unstable parts of the structure. You can make the loops a little more stable by adding G's to them. The procedure works as such, mouse over the loop and note the free energy of the loop. Then mutate each of the adjacent residues in the loop and see which improves (makes smaller or more negative) the free energy. Only mutate one at a time, hitting Undo (Z) after each. Once you determine which positions will be most adventagous, try and mutate as many as needed to minimize the free energy of the loop. Do not bother trying to mutate the nucleotides in the loop that are not adjacent to the stacks, they will have no effect.

4.
Checking for Mispairing. This is probably the most important and agrueably most difficult part of the process. The idea is to make each stretch of nucleotides in a stack have a unique sequence. This is where the Flip Nucleotide feature comes in handy. You have to look for sections of nucleotides that might mispair with the wrong nucleotides. Some easier things to look for include dinucleotide repeats (i.e. AUAUAU). This will likely be one of the biggest causes of failure for the RNA synthesis. While you are doing this, you may want to avoid trinucleotide repeats, because if you have a couple of them, they are more likely to mispair, and they are less stable

5.
Submit! Look over your RNA one last time and make sure that it looks good. Then click the submit button in the dock, give it an awesome name and send it on it's way.
Photo of Matt Baumgartner [mpb21]

Matt Baumgartner [mpb21], Alum

  • 128 Posts
  • 33 Reply Likes
  • happy

Posted 8 years ago

  • 16
Photo of dimension9

dimension9

  • 186 Posts
  • 45 Reply Likes
I thought this might be a good topic and an appropriate place to add a realization I just had while studying how round one of "The Cross" solutions came apart when switching to "Natural Mode."

One design in particular made me realize something I never thought to watch out for before, when creating a design: I was looking at Chesterfield's "thecross-1" ("the drunken ant"), and watching how it failed as it switched to natural mode...

What happened was that the left and right legs of the cross stayed together laterally, but moved apart in opposite directions longitudinally. Meanwhile, the top and bottom legs split apart laterally and collapsed toward one another into the center (like a dancer doing a "split" on a mirror). This meant that the inside left of the upper leg wound up kissing and bonding with the inside left of the lower leg, and, in mirror-fashion, the inside right of the upper leg wound up similarly bonded to the inside right of the lower leg.

I highly recommend watching this, as it was very enlightening.

Now, in my round 2 designs, (all made before this) I spent a significant amount of time trying to absolutely minimize any unintended bonding matches between the inner half-facing legs across the 90 degree angle spaces that separated them.

However, watching Chesterfield's design meltdown into a "dancer-split," made me realize that the insides of the legs may have been the vastly superior location to focus these efforts - and that indeed, the insides of a design in general, may be intrinsically very much more likely, or prone to bonding in wrong places than anywhere on the outside.

I have no idea if this is correct in other cases, but I thought it may be helpful enough in some instances to warrant this caution, and at least bring it to others' attention.

I will be watching for this kind of "inside-edge mis-bonding" very carefully in all my future designs.

I'd love to hear others; thoughts and impressions on this.

-d9
Photo of dimension9

dimension9

  • 186 Posts
  • 45 Reply Likes
Thought this might be a good place to link to this table:

http://getsatisfaction.com/eternagame...
Photo of markosys

markosys

  • 3 Posts
  • 2 Reply Likes
I've a few questions and suggestions centered around the RNA lab:

(1) When viewing a synthesized sequence in experimental mode, what do the gray nucleotides signify? Initially, I thought that they were gray because they were locked out from user changes, but in reviewing donald's leading Lab 2, Round 1 design, the gray nucleotides extend into the stack neck. The youtube tutorial doesn't mention them.

(2) I've used your written methods and suggestions above to strengthen and validate my designs and wondered if they and other fundamental validation/design rules could be made available as a RNA lab feature. Perhaps something that identified that additional loop stabilization was possible or that pointed out di/trinucleotide and other types of mispairings. I believe I came across some discussion of giving every submitted design a score which would be excellent. In the gaming spirit, I'd be happy to pay points that would lead the way to a stronger design (like hiring a Phd consultant).

(3) As an alternative or bridge while you're developing a validation/design rule suite :-), it would be useful if there were a method for exporting a design sequence so that the programming inclined among us could develop their own external validations. Scanning for mispairings late at night makes for tired eyes after a while.

(4) Would it be possible to make user-based solutions from #3 available to the community as a plug-in for the RNA lab review tool? Does your framework allow for extensions?
Photo of stansch

stansch

  • 9 Posts
  • 1 Reply Like
So OK, I got 10,000+ points by ding puzzles and now I am able to enter the RNA lab.
and do those configurations. I was able to make the bulging cross stable pretty quickly, but how do I know if it is a "good" design. What criteria should I use to vote for someone else's design?

This reply was created from a merged topic originally titled
Criteria for "good" design.
Photo of dimension9

dimension9

  • 186 Posts
  • 45 Reply Likes
Hi Stansch, start by looking up past lab design winners & studying them. Read as much as you can in Get-Satisfaction; there are a lot of topics discussing design; here is one: http://getsatisfaction.com/eternagame...
there are many more...
Photo of dimension9

dimension9

  • 186 Posts
  • 45 Reply Likes
Fix of Broken Link in fist paragraph of Post (to "The Strategy Guide to Solve EteRNA Puzzles):

http://getsatisfaction.com/eternagame...

-d9
Photo of Eli Fisker

Eli Fisker

  • 2219 Posts
  • 481 Reply Likes
Here is my strategy guide to the lab. It is far from complete, but it is meant as a help for the new players in our game.

My strategy guide to the lab
Photo of Eli Fisker

Eli Fisker

  • 2219 Posts
  • 481 Reply Likes
Actually I ended up making my whole profile page into a lab guide with extensions. So here is the link to all of it.

Have fun!
Photo of Adrien Treuille

Adrien Treuille, Alum

  • 243 Posts
  • 33 Reply Likes
So cool!
Photo of Eli Fisker

Eli Fisker

  • 2219 Posts
  • 481 Reply Likes
And as it appear the link to my page is broken, here it is again.

There is also an advice section on how to get started in lab in A Comprehensive Guide to EteRNA.
Photo of dyohn123

dyohn123

  • 2 Posts
  • 0 Reply Likes
Excellent information, thank you all for posting!
Photo of Adely Villarreal Ogburn

Adely Villarreal Ogburn

  • 1 Post
  • 0 Reply Likes
Can't wait to learn more from this and thanks bud
Photo of Kitty Rivershack

Kitty Rivershack

  • 1 Post
  • 1 Reply Like
I have looked at the guides and I am still baffled. Each of the cloud lab entries I've tried has a restriction such as "Must not have 4 or more consecutive Gs" and yet the RNA already has a string of 4 locked consecutive Gs! How is it possible to submit a usable solution when the given pattern violates the restrictions in the first place?

Head scratchin'
Photo of dreamdragon

dreamdragon

  • 1 Post
  • 0 Reply Likes
its broke
Photo of Omei Turnbull

Omei Turnbull, Player Developer

  • 962 Posts
  • 302 Reply Likes
Hi, dreamdragon. You need to earn your "Joined Lab" badge (10,000 coins) in order to get access to the lab part of Eterna.